Cross-lingual and cross-cultural generalization

Broaden the DNE-ElecDeb dataset and the Joint Argument--Entity Tagging framework beyond English-language U.S. presidential and vice-presidential debate transcripts to other languages and cultures.

Background

DNE-ElecDeb is constructed exclusively from English-language transcripts of U.S. presidential and vice-presidential debates. The authors caution that discourse patterns learned from this setting may not transfer reliably to other linguistic or cultural contexts, identifying expansion to additional languages as unresolved.

References

We acknowledge certain discourse patterns may not transfer to other languages or cultures, and therefore broadening the set of languages remains an open direction.

— Who Argues What? Joint Argument-Entity Detection and Classification in Political Debates  (2609.10192 - Cava et al., 9 Sep 2026) in Limitations, Language Use